Points – the mean raw score of individuals in each of the disciplines (full raw data cannot be shown for privacy reasons), against the proportion of men in the discipline (blue – men, red – women). …
Performance-Based Research Fund (PBRF) data using individuals at University of Canterbury (UC) (N=384). (A) Score against number of research outputs over the period of the PBRF assessment. Trend …
Scheme | PBRF | ARC | CIHR | EIGE | ||
---|---|---|---|---|---|---|
Country | Aotearoa New Zealand | Australia | Canada | Predominantly EU | ||
Time span | 2000–06 | 2007–12 | 2013–18 | 2010–19 | 2011–16 | 2019 |
Disciplines | 42 | 42 | 43 | 22 | 4 | 8 |
Evaluation | Research quality score (200–700) | Funding success | ||||
Women | 1708 | 2658 | 3297 | 46,231 | 8143 | 49,863 |
Men | 2522 | 4005 | 4181 | 130,331 | 15,775 | 85,305 |
Data points | 4230 | 6663 | 7487 | 440 | 16 | 333 |
Outcome | Mean score | Success rate | ||||
Women | 272 | 368 | 379 | 25.7% | 14.7% | 27.0% |
Men | 351 | 424 | 433 | 27.1% | 16.4% | 30.3% |
Wilkinson notation is used to indicate interaction terms.
Response | Maximal model | Best model (lowest AIC) | |
---|---|---|---|
PBRF | |||
ARC | |||
CIHR | |||
EIGE |
Mean (and standard deviation) of score and age at each time point. ***p<0.001, *0.01<p<0.05, two-sided t-test men/women.
Time period | 2006 | 2007–12 | 2013–18 | |
---|---|---|---|---|
Disciplines | 42 | 42 | 43 | |
N | Women | 1708 | 2658 | 3297 |
Men | 2522 | 4005 | 4181 | |
Mean score (std) | Women | 272 (161)*** | 368 (139)*** | 379 (142)*** |
Men | 351 (163) | 424 (143) | 433 (146) | |
Mean age (std) | Women | 44.5 (9.6)* | 47.8 (10.4)*** | 47.7 (11.1)*** |
Men | 45.3 (10.1) | 50.0 (10.9) | 49.5 (11.4) |
Table S1: Comparative descriptions of each of the four datasets.
Table S2: Summary statistics of the Performance-Based Research Fund (PBRF) data by discipline.
Includes the number of women and men, mean age, and mean score for each gender and proportion of men in each discipline.
Detailed model output for all Performance-Based Research Fund (PBRF), Australian Research Council (ARC), Canadian Institute of Health Research (CIHR), and European Institute of Gender Equality (EIGE) data candidate models.
Each sheet shows the coefficients and p-values for the full range of candidate models trialled for each dataset to predict either Score (PBRF) or Funding success (ARC, CIHR, EIGE). Coefficient names (first column) are given in Wilkinson’s notation. ΔAIC is the change in AIC between the best fit model (minimum AIC, ΔAIC = 0) and the candidate model. r-Squared is also given for the linear model (PRF only). Individual country coefficients and p-values for the EIGE data are not shown for clarity.
Detailed model output for the bibliometric models.
Sheet 1. Single bibliometric models .The coefficients and p-values for each of the five bibliometrics. ΔAIC is the change in AIC between the best fit model, with minimum AIC and ΔAIC = 0, and the candidate model. r-Squared is also given. Sheet 2. Two bibliometric model .The coefficients and p-values for each of the other four bibliometrics. ΔAIC is the change in AIC between the only model (minimum AIC, ΔAIC = 0) and the candidate model. r-Squared is also given. Sheet 3. Coefficients and p-values for .
DataARC.xlsx.
Field, Gender, Year, Number of applicants, Number of successes, proportion men (from external data).
DataCIHR.xlsx.
Field, Gender, Type of grant, Time period (before or after funding change), Number of applicants, number of successes, proportion male (from application numbers).
DataEIGE.xlsx.
Country, Field, Gender, Number of applicants, Number of successes, proportion male (from applicant data), proportion male from external data.